An integrative systematic revision and biogeography of Rhynchocalamus snakes (Reptilia, Colubridae) with a description of a new species from Israel

Abstract

Background: The colubrid snakes of the genus Rhynchocalamus are seldom studied and knowledge of their ecology and life history is scarce. Three species of Rhynchocalamus are currently recognized, R. satunini (from Turkey eastwards to Iran), R. arabicus (Yemen and Oman), and R. melanocephalus (from the Sinai Peninsula northwards to Turkey). All are slender, secretive, mainly nocturnal and rare fossorial snakes. This comprehensive study is the first to sample all known Rhynchocalamus species in order to review the intra-generic phylogenetic relationships and historical biogeography of the genus.

Methods: We revised the systematics of Rhynchocalamus using an integrative approach and evaluated its phylogeography. The phylogenetic position within the Colubridae and the phylogenetic relationships within the genus were inferred using 29 individuals belonging to the three known species, with additional sampling of two other closely-related genera, Muhtarophis and Lytorhynchus. We analysed three mitochondrial (12S, 16S, cytb) and one nuclear (c-mos) gene fragments. Phylogenetic trees were reconstructed using maximum likelihood and Bayesian inference methods; the latter method also used to provide the first time-calibrated molecular phylogeny of the genus. We generated a nuclear network and carried out a topology test and species delimitation analysis. Morphological comparisons were used to differentiate among species and to describe a new species from Israel. The studied material was comprised of 108 alcohol-preserved specimens, 15 photographs, and data from the literature for the examination of 17 mensural, 14 meristic, and two categorical characters.

Results: The molecular results support Rhynchocalamus as monophyletic, and as having split from its sister genus Lytorhynchus during the Late Oligocene. The three recognized species of Rhynchocalamus comprise four independently evolving groups. The molecular results reveal that the genus began to diverge during the Middle Miocene. We revealed that the best-studied species, R. melanocephalus, is paraphyletic. A population, formally ascribed to this species, from the Negev Mountain area in southern Israel is phylogenetically closer to R. arabicus from Oman than to the northern populations of the species from Israel, Syria and Turkey. Herein we describe this population as a new species: Rhynchocalamus dayanaesp. nov.

Discussion: We identify four species within Rhynchocalamus: R. satunini, R. arabicus, R. melanocephalus, and R. dayanaesp. nov., the latter, to the best of our knowledge, is endemic to southern Israel. The onset of Rhynchocalamus diversification is very old and estimated to have occurred during the Middle Miocene, possibly originating in the Levant region. Radiation probably resulted from vicariance and dispersal events caused by continuous geological instability, sea-level fluctuations and climatic changes within the Levant region.

Keywords: Arabia; Diversification; Evolution; Middle east; Phylogeny; Phylogeography; Reptiles; Taxonomy.

Figure 1. Distribution map of Rhynchocalamus.

Localities of the material analysed in this study, including type localities (star), samples used for the genetic analyses (circle), specimen vouchers for the morphological examinations (diamond) and photographic material (square). Numbers correspond to samples listed in Table S1 and colours to specimens in Fig. 2 and Figs. S1–S4. Taxon names correspond to changes proposed in this paper. Spatial data modified from various sources (IUCN-http://www.iucnredlist.org; Bar & Haimovitch, 2013; Sindaco, Venchi & Grieco, 2013).

Figure 2. Phylogenetic relationships within Rhynchocalamus.

(A) Maximum likelihood gene tree inferred from the concatenated dataset of the mitochondrial (12S, 16S, cytb) and nuclear (c-mos) gene fragments (dataset 3). Support values near the nodes indicate bootstrap and posterior probability (values ≥ 70%/ ≥ 0.95, respectively). Age estimates based on external calibration points of Hemorrhois and Hierophis subgroups (see ‘Materials and Methods’) are indicated near the relevant nodes and include the mean and, between parentheses, the 95% highest posterior densities (HPD) confidence interval. (B) Haplotype network of the c-mos nuclear marker. Circle size is proportional to the number of alleles. Sample codes and colours correlate to specimens in Table S1 and in Fig. 1 and Figs. S1–S4. Taxon names correspond to changes proposed in this paper.

Figure 3. Type series of Rhynchocalamus dayanae sp. nov.

Head and habitus (photos by Oz Rittner). (A, B) Holotype HUJ.R21704; (C, D) Paratype HUJ.R21705 (no skull); (E, F) Paratype TAU.R15930; (G, H) Paratype TAU.R17093. Scale bars = 2.5 mm.

Figure 4. Habitus comparisons of Rhynchocalamus taxa.

Dorsal view. (A) R. dayanae sp. nov. (unvouchered specimen; Road no. 40, near Mitzpe Ramon, Negev Mountain, Israel; photo by Simon Jamison); (B) R. melanocephalus (ZMHRU2007-69; Tartus, Syria; photo by Bayram Göçmen); (C) R. satunini (ZMHRU2015/0; Artuklu, Mardin province, Turkey; photo by Bayram Göçmen); (D) R. arabicus (CN4780; Wadi Ayoun, Dhofar Governorate, Oman; photo by Gabriel Martínez).

Figure 5. Head comparisons of Rhynchocalamus taxa.

Lateral and dorsal views. (A, B) R. dayanae sp. nov. (Holotype, HUJ.R21704; photos by Oz Rittner); (C, D) R. melanocephalus (Negev population; HUJ.R22058; photos by Oz Rittner); (E, F) R. melanocephalus (Mediterranean ecoregion population; BMNH1946.1.3.29; photos by Salvador Carranza); (G, H) R. arabicus (CN4780; photos by Gabriel Martínez); (I, J) R. satunini (ZISP17098; photos by Daniel Melnikov).

Figure 6. Habitat of R. dayanae sp. nov

The Negev Mountain region, road no. 171, Israel (the general area where the type material was collected).